A New Nearest Centroid Neighbor Classifier Based on K Local Means Using Harmonic Mean Distance

Mehta, Sumet; Shen, Xiang-Jun; Gou, Jiangping; Niu, Dejiao

doi:10.3390/info9090234

Cited by 30 publications

(13 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The aforementioned NPE is based on graph embedding and the neighbor relationship is measured by an artificially-constructed adjacent graph. Usually the KNN [2] criteria is used to construct an adjacent graph which makes the performance of the NPE very sensitive to the parameter of the neighborhood size. Thus, to overcome the neighborhood size sensitivity in NPE, we are motivated to propose a novel DR technique called weighted neighborhood preserving ensemble embedding (WNPEE).…”

Section: The Proposed Wnpee Methodsmentioning

confidence: 99%

“…In many machine learning applications, such as data classification [1,2], face recognition [3,4], signal processing [5,6], and text categorization [7,8], the input data is usually high-dimensional which makes the calculations too complex, as well as requiring more computational time. In order to boost the performance and computational efficiency, various dimensionality reduction (DR) techniques have been proposed to preprocess these high-dimensional data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Weighted Neighborhood Preserving Ensemble Embedding

2019

Self Cite

View full text Add to dashboard Cite

Neighborhood preserving embedding (NPE) is a classical and very promising supervised dimensional reduction (DR) technique based on a linear graph, which preserves the local neighborhood relations of the data points. However, NPE uses the K nearest neighbor (KNN) criteria for constructing an adjacent graph which makes it more sensitive to neighborhood size. In this article, we propose a novel DR method called weighted neighborhood preserving ensemble embedding (WNPEE). Unlike NPE, the proposed WNPEE constructs an ensemble of adjacent graphs with the number of nearest neighbors varying. With this graph ensemble building, WNPEE can obtain the low-dimensional projections with optimal embedded graph pursuing in a joint optimization manner. WNPEE can be applied in many machine learning fields, such as object recognition, data classification, signal processing, text categorization, and various deep learning tasks. Extensive experiments on Olivetti Research Laboratory (ORL), Georgia Tech, Carnegie Mellon University-Pose and Illumination Images (CMU PIE) and Yale, four face databases demonstrate that WNPEE achieves a competitive and better recognition rate than NPE and other comparative DR methods. Additionally, the proposed WNPEE achieves much lower sensitivity to the neighborhood size parameter as compared to the traditional NPE method while preserving more of the local manifold structure of the high-dimensional data.

show abstract

Section: The Proposed Wnpee Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Weighted Neighborhood Preserving Ensemble Embedding

2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…Figure illustrates this idea for the three actuation phases. Specifically, for a given structural state, the harmonic mean of the distances between the centroids

{scriptY}_{l}^{φ}

and the map point

{boldy}_{φ}^{21}

for all the actuation phases

φ = 1, \dots, 4

\frac{1}{\frac{1}{4} {false\sum}_{φ = 1}^{4} \frac{1}{‖ Y_{l}^{φ} - y_{φ}^{21} ‖_{2}}} .

Therefore,

l = \arg \max_{l = 1, \dots, E} \sum_{φ = 1}^{4} \frac{1}{‖ Y_{l}^{φ} - y_{φ}^{21} ‖_{2}} = \arg \min_{l = 1, \dots, E} \frac{1}{\frac{1}{4} {false\sum}_{φ = 1}^{4} \frac{1}{‖ Y_{l}^{φ} - y_{φ}^{21} ‖_{2}}} .

Mehta et al also use the harmonic distance to define a pattern classification technique similar to the

k

‐nearest neighbor classifier.…”

Section: Case Studies: Aluminum Plate With Four Pztsmentioning

confidence: 99%

Vibration‐based detection and classification of structural changes using principal component analysis and ‐distributed stochastic neighbor embedding

Agis

Burgos

Pozo

2020

Struct Control Health Monit

View full text Add to dashboard Cite

Summary This paper describes a structural health monitoring strategy to detect and classify structural changes in structures that can be equipped with sensors. The proposed approach is based on the t‐distributed stochastic neighbor embedding ( t‐SNE), a nonlinear technique that can represent the local structure of high‐dimensional data collected from multiple sensors in a plane or spatial representation. We propose the following basic steps for the detection and classification. First, the raw data are preprocessed: We scale the data using the mean‐centered group scaling and apply principal component analysis to reduce the dimensionality of the scaled data. Second, t‐SNE is applied to represent the scaled and reduced data as points in a plane, defining a cluster for each structural state. Finally, the current structure to be diagnosed is associated with a cluster (or structural state) using three different strategies: (a) the smallest point‐centroid distance; (b) the majority voting; and (c) the sum of the inverse distances. The combination of t‐SNE with our preprocessing and the three proposed classification strategies significantly improves the quality of the clusters that represent different structural states. We evaluate the performance of our method using experimental data from an aluminum plate instrumented with piezoelectric transducers. Results are presented in the time domain, and they reveal the high classification accuracy and strong performance of this method, with a percentage of correct decisions close to 100% in several scenarios.

show abstract

“…In addition, the method does not perform well in certain data sets, especially data sets with outlier [10]. Outlier is an observation result that is different in nature with a majority of other observations [11].…”

Section: Introductionmentioning

confidence: 97%

Nearest Centroid Classifier with Outlier Removal for Classification

Bawono

Bahtiar

Supianto

2020

JITeCS

View full text Add to dashboard Cite

Classification is a part of data mining which is very useful for many purpose. There are so many method to classify, however a classification process need good source of data to deliver good performance. In the real world, not all data is good. Some of the data contains noise and outlier that can reduce classification performance. However, there are few research of classification with an outlier removal process, especially for Nearest Centroid Classifier Method. Nearest Centroid Classifier is a good yet simple classification method. However, Nearest Centroid Classifier suffers performance loss like the other classification method. The proposed methodology consists of two stages. First, preprocess the data with outlier removal, removes points which are far from the corresponding centroids, the center point of a class. Second, classify the outlier removed data or cleaned data. The experiment covers six data sets which have different characteristic. The results indicate that outlier removal as preprocessing method provide better result by 0.76% to 3.16% for improving Nearest Centroid Classifier performance on most data set.

show abstract

A New Nearest Centroid Neighbor Classifier Based on K Local Means Using Harmonic Mean Distance

Cited by 30 publications

References 33 publications

Weighted Neighborhood Preserving Ensemble Embedding

Weighted Neighborhood Preserving Ensemble Embedding

Vibration‐based detection and classification of structural changes using principal component analysis and ‐distributed stochastic neighbor embedding

Nearest Centroid Classifier with Outlier Removal for Classification

Contact Info

Product

Resources

About